The identification schema for drugs proposed by this invention extends the concepts described in U.S. Pat. Nos. 5,031,937 and 5,845,264 of Dr. Gerhard Nellhaus and in particular the system for bar coding drug tablets, capsules, pills and the like for identification.
In selecting a bar code scheme for drugs, ease of reading both by machine and by visual inspection is an important criterion. A two dimensional bar code of the data matrix type is compact and can be easily scanned by an image capture device. Certain sweep scanners can read primitive data matrix type codes. However, the use of charged, coupled devices (CCD""s) for images capture and software for image rotation and decoding makes the matrix pattern more practical for small objects.
For the application of identifying drugs, simplicity is desired. A 4xc3x974 matrix is compact and lends itself to code in hexadecimal. Any standard reference can decode the hexadecimal to obtain alphanumeric characters Theoretically, a 4xc3x974 matrix can code 65,536 items using a binary half-bite system.
However, orientation is a problem for a correct read of a matrix symbol, which may be a square mark at the center of a round pill. For most applications in drug marking, a 5xc3x975 xe2x80x9cadaptedxe2x80x9d matrix or a 6xc3x976 matrix using Data Matrix(trademark) alignment protocol is appropriate. For the very small pills a 4xc3x974 matrix can be used with three of the corner squares having a common marking, i.e. black with the remaining square unmarked. This allows a definition of orientation, but reduces the date bits to 12 for itemizing only 4,096 objects.
Therefore, in the preferred bar code schema for imprinted drugs, that minimally satisfies the current FDA requirement for imprinting solid form drugs, a modified 4xc3x974 matrix data field is used. To accommodate more than 4,096 objects, except for the reserve code for ultra small drugs, a 5xc3x975 matrix is the preferred minimum.
Alternately, a matrix outline and a header bar for orientation can be used.
This results in a 6xc3x974 matrix, since a space column should be incorporated between the data matrix and the header bar. Notably, the header bar could constitute the alphanumeric, advised, but not required by the FDA with little useful purpose other than orientation.
The 5xc3x975 matrix, however, advantageously allows a solid header bar to be combined with an alternating checkerboard square pattern to both define the matrix size and allow for determination of granularity, which is an aid in reading. The granularity allows the size of the data bits to be determined.
As a last resort, the full protocol of the Data Matrix(trademark) system can be implemented with a 6xc3x976 matrix, although ironically this is too small to comply with the current ANSI Data Matrix(trademark) standards which start with a 12xc3x9712 matrix.
Larger matrices are not required unless larger numbers of drugs are to be marked. Since marked information is desired to be informative, encryption codes are not required, thereby maximizing the data content in drug identification applications using a data matrix type protocol.
One solution to the competing interests of maximizing the visibility of the data squares within a data matrix orientation and granularity frame, and minimizing the overall size of the matrix is provided by doubling the size of the data squares in comparison with the squares forming the frame. In this manner, a 4xc3x974 matrix of double size data squares in a frame of unitary size squares forms an effective 10xc3x9710 matrix. A 5xc3x975 matrix of double size data squares with an orientation and granularity frame of unitary size squares effectively forms a 12xc3x9712 matrix, coincidentally the minimum matrix for an ANSI Data Matrix(trademark) symbol. It is to be understood, however, that the error correction and validation coding within the matrix of a Data Matrix(trademark) symbol are not employed. In the hybrid Data Matrix(trademark) system disclosed, the increased size of the data squares substantially improves the likelihood of an accurate reading, and, more than off-sets the loss of error correction coding within the data field.
It is preferred that the matrix of the data matrix imprint be contained within a perimeter graphic that is useful for visual determination or differentiation of the drug by the patient. This graphic may be used in combination with other alphanumeric or symbolic marks, or with the pill shape or color for visual identification. The preferred graphic ideograph or icon should be easily recognizable for categorical identification. Additionally, the ideograph or icon should be of a type that can be easily communicated orally, for example, from one person to another by telephone, to aid in determining the pill""s identity.
Upon implementation of the coding schema, a code reader pen is used for automatic identification of the drug item without resorting to tables and directories. The code reader pen is a self contained instrument similar in operation to a push-top, ball point pen with a light source, and a data retrieval lens that focuses the code imprint on a charged, coupled device (CCD) to capture a readable image. The pen includes a small processor and memory to decode the pattern, for example, by rotation and pattern matching against a set of stored templates or patterns. A database listing of drug names, for example, is displayed in an elongated display along the side of the pen which the user reads, like a thermometer, when making an identification using the code reader pen.